Language Processing Strategies Research Articles

Leveraging Machine Learning and Patient Reviews for Developing a Drug Recommendation System to Reduce Medical Errors

Background In the rapidly evolving pharmaceutical industry, drug efficacy and safety stand as critical concerns. The vast accumulation of data, including customer feedback, drug popularity, and usage details, offers a rich resource for improving healthcare outcomes. Aims The primary aim of this study is to harness machine learning and Natural Language Processing (NLP) techniques to sift through extensive pharmaceutical data, identifying the most effective drugs for various conditions and uncovering patterns that could guide better decision-making in drug efficacy and safety. Objective This research seeks to construct a sophisticated model capable of analyzing diverse data points to pinpoint the most efficacious drugs for specific health conditions, thereby providing pharmaceutical companies with data-driven insights to optimize drug safety and effectiveness. Methods Employing a blend of Natural Language Processing (NLP) and machine learning strategies, the study analyzes a comprehensive dataset featuring customer reviews, drug popularity metrics, usage information, and other relevant data collected over an extended period. This methodological approach aims to reveal latent trends and patterns that are crucial for assessing drug performance. Results The developed model adeptly identifies leading medications for various conditions, elucidating efficacy and safety profiles derived from patient reviews and drug utilization trends. These findings furnish pharmaceutical companies with actionable intelligence for enhancing drug development and patient care strategies. Conclusion The integration of machine learning and NLP for the analysis of vast drug-related datasets presents a powerful method for advancing drug efficacy and safety. The insights yielded by the proposed model significantly empower the decision-making processes of the pharmaceutical industry, ultimately fostering improved health outcomes for patients.

Natural Language Understanding for Navigation of Service Robots in Low-Resource Domains and Languages: Scenarios in Spanish and Nahuatl

Human–robot interaction is becoming increasingly common to perform useful tasks in everyday life. From the human–machine communication perspective, achieving effective interaction in natural language is one challenge. To address it, natural language processing strategies have recently been used, commonly following a supervised machine learning framework. In this context, most approaches rely on the use of linguistic resources (e.g., taggers or embeddings), including training corpora. Unfortunately, such resources are scarce for some languages in specific domains, increasing the complexity of solution approaches. Motivated by these challenges, this paper explores deep learning methods for understanding natural language commands emitted to service robots that guide their movements in low-resource scenarios, defined by the use of Spanish and Nahuatl languages, for which linguistic resources are scarcely unavailable for this specific task. Particularly, we applied natural language understanding (NLU) techniques using deep neural networks and transformers-based models. As part of the research methodology, we introduced a labeled dataset of movement commands in the mentioned languages. The results show that models based on transformers work well to recognize commands (intent classification task) and their parameters (e.g., quantities and movement units) in Spanish, achieving a performance of 98.70% (accuracy) and 96.96% (F1) for the intent classification and slot-filling tasks, respectively). In Nahuatl, the best performance obtained was 93.5% (accuracy) and 88.57% (F1) in these tasks, respectively. In general, this study shows that robot movements can be guided in natural language through machine learning models using neural models and cross-lingual transfer strategies, even in low-resource scenarios.

Automatically Generated Datasets: Present and Potential Self-Cleaning Coating Materials

The rise of urbanization coupled with pollution has highlighted the importance of outdoor self-cleaning coatings. These revolutionary coatings contribute to the longevity of various surfaces and reduce maintenance costs for a wide range of applications. Despite ongoing research to develop efficient and durable self-cleaning coatings, adopting systematic research methodologies could accelerate these advancements. In this work, we use Natural Language Processing (NLP) strategies to generate open- and traceable-sourced datasets about self-cleaning coating materials from 39,011 multi-disciplinary papers. The data are from function-based and property-based corpora for self-cleaning purposes. These datasets are presented in four different formats for diverse uses or combined uses: material frequency statistics, material dictionary, measurement value datasets for self-cleaning-related properties and optical properties, and sentiment statistics of material stability and durability. This provides a literature-based data resource for the development of self-cleaning coatings and also offers potential pathways for material discovery and prediction by machine learning.

Shielding Online Communities: Natural Language Processing and Machine Learning Strategies against Social Media Intimidation

With the usage of the internet and the growing prominence of communities, like social media we have witnessed a rise in cybercrime. Among these crimes one that stands out is Intimidator, which affects both people and adults alike. The increasing incidents of cyberbullying have led to consequences such as anxiety, aggression, depression and tragically even suicide. Consequently, there is now a pressing need for content regulation on social media platforms. This research focuses on developing a model of identifying text-based bullying messages and comments by categorizing them into five distinct types; Violence, Vulgar language Offensive content, sexually explicit material, and Hate Speech. The proposed approach involves utilizing Natural Language Processing (NLP) techniques with Machine Learning methods. The dataset is initially. Processed to remove information before extracting meaningful features. Finally, the model undergoes training and testing to ensure reliable results, in detecting instances of Intimidator in text-based data.

Frequent native language use supports phonemic and semantic verbal fluency in L1 and L2: An extended analysis of verbal fluency task performance in an L1 language attrition population

Purpose: Verbal fluency (VF) tasks are an efficient tool for exploring lexical retrieval. Attriters generally produce fewer words in semantic categories than monolinguals, but the processes underlying attriters’ lexical retrieval remain unclear. Phonemic tasks are scarcely applied in language attrition studies. We aim to identify processes underlying attriters’ phonemic verbal fluency (PVF) and semantic verbal fluency (SVF) performance by extending the analysis beyond total scores in the first (L1) and second (L2) language. Design: We modeled total scores and temporal aspects of task performance in L1 and L2 for the attriters, with L1 use and length of residence (LoR) as predictors. We analyzed the number and types of errors between languages and compared L1 task performance with a matched monolingual group. Data and analysis: Attriters’ ( N = 38) phonemic and semantic task performance in L1 (Finnish) and L2 (English) were modeled, and results were contrasted to a matched monolingual control group ( N = 50). Findings: Attriters demonstrated strong proficiency in L1 and similar lexical retrieval processes in L1 and L2 after 20 years of immersive L2 exposure. Frequent L1 use supported overall performance but slowed performance down in both languages. Compared with monolinguals, attriters show a disadvantage in SVF, but not due to slower initiation. Instead, attriters rely more on rapid L1 retrieval than monolinguals. Originality: Prior research on VF tasks in language attrition populations has mainly focused on SVF total scores in L1. Our investigation explores underlying mechanisms of L1 and L2 task performance in PVF and SVF within a language attrition population and contrasts L1 performance to a matched monolingual group performance. Implications: This study demonstrates the importance of investigating L2 alongside L1 in language attrition studies for a holistic approach to language processing strategies. Our findings enhance understanding of the processes underlying VF task performance, emphasizing the significance of VF task early-stage performance.

Evolving natural language processing towards a subjectivist inductive paradigm.

Gin reflects on the advantages of developing new natural language processing strategies that invite both researchers and their AI “assistants” into a collaborative and transparent process of inquiry.

Linguistic landscaping: Eyetracking multilingual information processing

This paper presents an eyetracking study on language reception in a multilingual society. It focuses on the reading behavior of 23 trilingual participants from South Tyrol, Italy, who speak and understand Italian, German and Ladin, the latter a minority language in that area. The eyetracking data show whether all three languages displayed on informational signs are read and processed, and whether processing depends on specific influential factors dependent of the participants or presentation of the language versions. This type of eyetracking application poses several methodological issues which are discussed in the paper. The eyetracking data are triangulated with data from a speech production task and a self-assessment of language skills. The initial results, show that Ladin is preferred over German and Italian in terms of overall fixation duration, that the order of processing follows the marked position top to bottom and left to right, and that participant's language used to describe the processed signs matches that of the language with the most attention per stimulus. Concerning specific language processing strategies initial interpretations are given which must be followed up on.

Additions of Conjunctive Vav (“and”) in 1QIsaa: Evidence for the Role of Language Processing Strategies

Abstract This article identifies one type of conjunctive vav (“and”) addition in 1QIsaa that accounts for over one-third of the 241 vav additions in the scroll. In these cases, vav serves to resolve structurally ambiguous verses in the biblical text of Isaiah. In 84 percent of the cases, vav resolves the ambiguous verses in a manner that accords with well-known language processing strategies. It is therefore argued that processing strategies were operative during the copying of 1QIsaa. Notably, more than half of vav additions that accord with processing strategies result in a verse division that differs from the MT’s reading tradition as recorded in its cantillation system. These findings offer a new perspective on the readings that are generated by vav additions when compared to the MT and also allow us to predict in which contexts this type of vav is likely to be found in other unpunctuated ancient texts.

Natural Language Processing for Virtual Reference Analysis

Objective – Chat transcript analysis can illuminate user needs by identifying common question topics, but traditional hand coding methods for topic analysis are time-consuming and poorly suited to large datasets. The research team explored the viability of automatic and natural language processing (NLP) strategies to perform rapid topic analysis on a large dataset of transcripts from a consortial chat service. Methods – The research team developed a toolchain for data processing and analysis, which incorporated targeted searching for query terms using regular expressions and natural language processing using the Python spaCy library for automatic topic analysis. Processed data was exported to Tableau for visualization. Results were compared to hand-coded data to test the accuracy of conclusions. Results – The processed data provided insights about the volume of chats originating from each participating library, the proportion of chats answered by operator groups for each library, and the percentage of chats answered by different staff types. The data also captured the top referring URLs for the service, course codes and file extensions mentioned, and query hits. Natural language processing revealed that the most common topics were related to citation, subscription databases, and finding full-text articles, which aligns with common question types identified in hand-coded transcripts. Conclusion – Compared to hand coding, automatic and NLP processing approaches have benefits in terms of the volume of data that can be analyzed and the time frame required for analysis, but they come with a trade-off in accuracy, such as false hits. Therefore, computational approaches should be used to supplement traditional hand coding methods. As NLP becomes more accurate, approaches such as these may widen avenues of insight into virtual reference and patron needs.

Anomaly Detection on Natural Language Processing to Improve Predictions on Tourist Preferences

Argumentation-based dialogue models have shown to be appropriate for decision contexts in which it is intended to overcome the lack of interaction between decision-makers, either because they are dispersed, they are too many, or they are simply not even known. However, to support decision processes with argumentation-based dialogue models, it is necessary to have knowledge of certain aspects that are specific to each decision-maker, such as preferences, interests, and limitations, among others. Failure to obtain this knowledge could ruin the model’s success. In this work, we sought to facilitate the information acquisition process by studying strategies to automatically predict the tourists’ preferences (ratings) in relation to points of interest based on their reviews. We explored different Machine Learning methods to predict users’ ratings. We used Natural Language Processing strategies to predict whether a review is positive or negative and the rating assigned by users on a scale of 1 to 5. We then applied supervised methods such as Logistic Regression, Random Forest, Decision Trees, K-Nearest Neighbors, and Recurrent Neural Networks to determine whether a tourist likes/dislikes a given point of interest. We also used a distinctive approach in this field through unsupervised techniques for anomaly detection problems. The goal was to improve the supervised model in identifying only those tourists who truly like or dislike a particular point of interest, in which the main objective is not to identify everyone, but fundamentally not to fail those who are identified in those conditions. The experiments carried out showed that the developed models could predict with high accuracy whether a review is positive or negative but have some difficulty in accurately predicting the rating assigned by users. Unsupervised method Local Outlier Factor improved the results, reducing Logistic Regression false positives with an associated cost of increasing false negatives.

FADOHS: Framework for Detection and Integration of Unstructured Data of Hate Speech on Facebook Using Sentiment and Emotion Analysis

Hate speech is a form of expression that assaults a person or a community based on race, origin, religion, sexual orientation, or other attributes. Although it can be expressed in multiple ways, both online and offline, the increasing popularity of social media has exponentially increased both its use and severity. Therefore, the aim of this research is to locate and analyze the unstructured data of selected social media posts that intend to spread hate in the comment sections. To address this issue, we propose a novel framework called FADOHS, which combines data analysis and natural language processing strategies, to sensitize all social media providers to the pervasiveness of hate on social media. Specifically, we use sentiment and emotion analysis algorithms to analyze recent posts and comments on these pages. Posts suspected of containing dehumanizing words will be processed before fed to the clustering algorithm for further evaluation. According to the experimental results, the proposed FADOHS framework is able to surpass the state-of-the-art approach in terms of precision, recall, and F1 scores by approximately 10%.

Alzheimer’s disease prevention: Within reach by 2025

AbstractBackgroundAlzheimer’s disease (AD) is a progressive neurodegenerative disease that affects memory and cognition. Symptoms and onset of AD occur up to 20 years after potential alterations in biology, allowing for a tremendous potential to intervene in disease progression to change the AD course. In 2015, the National Plan to Address Alzheimer Disease as part of the National Alzheimer’s Project Act was signed into law which proposed to effectively prevent AD by 2025 sparking new interested in the clinical and research field to achieve these goals.MethodWe used a novel combination of text‐mining and natural language processing strategies to identify (i) AD risk factors, (ii) therapeutics that can target risk factor pathways and (iii) studies supporting therapeutics in the currently published PubMed database. To classify the literature relevant to AD preventive strategies, we developed a relevance score based on STRING score for protein‐protein interactions and a confidence score on Bayesian inference. This led to generation of a ranked list of candidate therapeutics to reduce AD risk.ResultBased on our strategy, 364 AD risk factors were identified mining 9,625 publications. Next, using drug databases, 694 FDA‐approved drugs were selected based on drug indications with a given risk factor. The computation of ranking scores enabled exclusion of publications that did not meet inclusion criteria and rank 45 drugs associated with reduced risk. Within this list, 22 therapeutics had at least one clinical study supporting AD risk reduction. The top 20 therapeutics included drugs within the categories of lipid‐lowering, anti‐inflammatory, anxiety/psychiatric, and metabolic‐related therapeutics.ConclusionOutcomes of our novel bioinformatic strategy supports therapeutic targeting of biological mechanisms and pathways underlying AD risk factors. Based on our analyses, we propose that early interventions that target pathways associated with increased risk of AD has the potential to achieve the goal of effectively preventing AD by 2025.

SENTIMENT ANALYSIS FOR SARCASTIC MESSAGES IN SOCIAL MEDIA USING DEEP LEARNING TECHNIQUES - AN EMPIRICAL STUDY

Sentiments is the term of opinion or views about any topic expressed by the people through a source of communication. Nowadays social media is an effective platform for people to communicate and it generates huge amount of unstructured details every day. It is essential for any business organization in the current era to process and analyse the sentiments by using machine learning and Natural Language Processing (NLP) strategies. Even though in recent times the deep learning strategies are becoming more familiar due to higher capabilities of performance. This paper represents an empirical study of an application of deep learning techniques in Sentiment Analysis (SA) for sarcastic messages and their increasing scope in real time. Taxonomy of the sentiment analysis in recent times and their key terms are also been highlighted in the manuscript. The survey concludes the recent datasets considered, their key contributions and the performance of deep learning model applied with its primary purpose like sarcasm detection in order to describe the efficiency of deep learning frameworks in the domain of sentimental analysis.

Preventing Alzheimer's disease within reach by 2025: Targeted-risk-AD-prevention (TRAP) strategy.

IntroductionAlzheimer's disease (AD) is a progressive neurodegenerative disease that currently affects 6.2 million people in the United States and is projected to impact 12.7 million worldwide in 2050 with no effective disease‐modifying therapeutic or cure. In 2011 as part of the National Alzheimer's Project Act, the National Plan to Address Alzheimer's Disease was signed into law which proposed to effectively prevent AD by 2025, which is rapidly approaching. The preclinical phase of AD can begin 20 years prior to diagnosis, which provides an extended window for preventive measures that would exert a transformative impact on incidence and prevalence of AD.MethodsA novel combination of text‐mining and natural language processing strategies to identify (1) AD risk factors, (2) therapeutics that can target risk factor pathways, and (3) studies supporting therapeutics in the PubMed database was conducted. To classify the literature relevant to AD preventive strategies, a relevance score (RS) based on STRING (search tool for the retrieval of interacting genes/proteins) score for protein–protein interactions and a confidence score (CS) on Bayesian inference were developed. To address mechanism of action, network analysis of protein targets for effective drugs was conducted. Collectively, the analytic approach, referred to as a targeted‐risk‐AD‐prevention (TRAP) strategy, led to a ranked list of candidate therapeutics to reduce AD risk.ResultsBased on TRAP mining of 9625 publications, 364 AD risk factors were identified. Based on risk factor indications, 629 Food and Drug Administration‐approved drugs were identified. Computation of ranking scores enabled identification of 46 relevant high confidence (RS & CS > 0.7) drugs associated with reduced AD risk. Within these candidate therapeutics, 16 had more than one clinical study supporting AD risk reduction. Top‐ranked therapeutics with high confidence emerged within lipid‐lowering, anti‐inflammatory, hormone, and metabolic‐related drug classes.DiscussionOutcomes of our novel bioinformatic strategy support therapeutic targeting of biological mechanisms and pathways underlying relevant AD risk factors with high confidence. Early interventions that target pathways associated with increased risk of AD have the potential to support the goal of effectively preventing AD by 2025.

How does nursing research differ internationally? A bibliometric analysis of six countries.

International nursing research comparisons can give a new perspective on a nation's output by identifying strengths and weaknesses. This article compares strengths in nursing research between six mainly English-speaking nations (Australia, Canada, Ireland, New Zealand, United Kingdom and United States). Journal authorship (percentage of first authorship by nationality) and article keywords were compared for Scopus-indexed journal articles 2008-2018. Three natural language processing strategies were assessed for identifying statistically significant international differences in the use of keywords or phrases. Journal author nationality was not a good indicator of international differences in research specialisms, but keyword and phrase differences were more promising especially if both are used. For this, the part of speech tagging and lemmatisation text processing strategies were helpful but not named entity recognition. The results highlight aspects of nursing research that were absent in some countries, such as papers about nursing administration and management. Researchers outside the United States should consider the importance of researching specific patient groups, diseases, treatments, skills, research methods and social perspectives for unresearched gaps with national relevance. From a methods perspective, keyword and phrase differences are useful to reveal international differences in nursing research topics.

SAT-LB111 Improving Classification of Diabetes Etiology in Electronic Resources Using Phenotype Algorithms and Polygenic Risk Scores

Electronic Health Records (EHR) contain rich data to identify and study diabetes. Many phenotype algorithms have been developed to identify research subjects with type 2 diabetes (T2D), but very few accurately identify type 1 diabetes (T1D) cases or more rare forms of monogenic and atypical metabolic presentations. Polygenetic risk scores (PRS) quantify risk of a disease using common genomic variants well for both T1D and T2D. In this study, we apply validated phenotyping algorithms to EHRs linked to a genomic biobank to understand the independent contribution of PRS to classification of diabetes etiology and generate additional novel markers to distinguish subtypes of diabetes in EHR data. Using a de-identified mirror of medical center’s electronic health record, we applied published algorithms for T1D and T2D to identify cases, and used natural language processing and chart review strategies to identify cases of maturity onset diabetes of the young (MODY) and other more rare presentations. This novel approach included additional data types such as medication sequencing, ratio and temporality of insulin and non-insulin agents, clinical genetic testing, and ratios of diagnostic codes. Chart review was performed to validate etiology. To calculate PRS, we used genome wide genotyping from our BioBank, the de-identified biobank linking EHR to genomic data using coefficients of 65 published T1D SNPS and 76,996 T2D SNPS using PLINK in Caucasian subjects. In the dataset, we identified 82,238 cases of T2D but only 130 cases of T1D using the most cited published algorithms. Adding novel structured elements and natural language processing identified an additional 138 cases of T1D and distinguished 354 cases as MODY. Among over 90,000 subjects with genotyping data available, we included 72,624 Caucasian subjects since PRS coefficients were generated in Caucasian cohorts. Among those subjects, 248, 6,488, and 21 subjects were identified as T1D, T2D, and MODY subjects respectively in our final PRS cohort. The T1D PRS did significantly discriminate well between cases and controls (Mann-Whitney p-value is 3.4 e-17). The PRS for T2D did not significantly discriminate between cases and controls using published algorithms. The atypical case count was too low to calculate PRS discrimination. Calculation of the PRS score was limited by quality inclusion of variants available, and discrimination may improve in larger data sets. Additionally, blinded physician case review is ongoing to validate the novel classification scheme and provide a gold standard for machine learning approaches that can be applied in validation sets.

Language comprehension in the social brain: Electrophysiological brain signals of social presence effects during syntactic and semantic sentence processing

Although, evolutionarily, language emerged predominantly for social purposes, much has yet to be uncovered regarding how language processing is affected by social context. Social presence research studies the ways in which the presence of a conspecific affects processing, but has yet to be thoroughly applied to language processes. The principal aim of this study was to see how syntactic and semantic language processing might be subject to mere social presence effects by studying Event-Related brain Potentials (ERP). In a sentence correctness task, participants read sentences with a semantic or syntactic anomaly while being either alone or in the mere presence of a confederate. Compared to the alone condition, the presence condition was associated with an enhanced N400 component and a more centro-posterior LAN component (interpreted as an N400). The results seem to imply a boosting of heuristic language processing strategies, proper of lexico-semantic operations, which actually entails a shift in the strategy to process morphosyntactic violations, typically based on algorithmic or rule-based strategies. The effects cannot be related to increased arousal levels. The apparent enhancement of the activity in the precuneus while in presence of another person suggests that the effects conceivably relate to social cognitive and attentional factors. The present results suggest that understanding language comprehension would not be complete without considering the impact of social presence effects, inherent to the most natural and fundamental communicative scenarios.

RegulonDB v 10.5: tackling challenges to unify classic and high throughput knowledge of gene regulation in E. coli K-12.

RegulonDB, first published 20 years ago, is a comprehensive electronic resource about regulation of transcription initiation of Escherichia coli K-12 with decades of knowledge from classic molecular biology experiments, and recently also from high-throughput genomic methodologies. We curated the literature to keep RegulonDB up to date, and initiated curation of ChIP and gSELEX experiments. We estimate that current knowledge describes between 10% and 30% of the expected total number of transcription factor- gene regulatory interactions in E. coli. RegulonDB provides datasets for interactions for which there is no evidence that they affect expression, as well as expression datasets. We developed a proof of concept pipeline to merge binding and expression evidence to identify regulatory interactions. These datasets can be visualized in the RegulonDB JBrowse. We developed the Microbial Conditions Ontology with a controlled vocabulary for the minimal properties to reproduce an experiment, which contributes to integrate data from high throughput and classic literature. At a higher level of integration, we report Genetic Sensory-Response Units for 200 transcription factors, including their regulation at the metabolic level, and include summaries for 70 of them. Finally, we summarize our research with Natural language processing strategies to enhance our biocuration work.

Identifying Suicide Ideation and Suicidal Attempts in a Psychiatric Clinical Research Database using Natural Language Processing

Research into suicide prevention has been hampered by methodological limitations such as low sample size and recall bias. Recently, Natural Language Processing (NLP) strategies have been used with Electronic Health Records to increase information extraction from free text notes as well as structured fields concerning suicidality and this allows access to much larger cohorts than previously possible. This paper presents two novel NLP approaches – a rule-based approach to classify the presence of suicide ideation and a hybrid machine learning and rule-based approach to identify suicide attempts in a psychiatric clinical database. Good performance of the two classifiers in the evaluation study suggest they can be used to accurately detect mentions of suicide ideation and attempt within free-text documents in this psychiatric database. The novelty of the two approaches lies in the malleability of each classifier if a need to refine performance, or meet alternate classification requirements arises. The algorithms can also be adapted to fit infrastructures of other clinical datasets given sufficient clinical recording practice knowledge, without dependency on medical codes or additional data extraction of known risk factors to predict suicidal behaviour.

Repetition priming reveals hemispheric differences in compound word processing

ABSTRACTThe left (LH) and right (RH) hemispheres are thought to implement different mechanisms for visual word recognition; the LH’s parallel encoding strategy is more efficient than the RH’s serial, letter-by-letter analysis. Here we examine differences in hemispheric language processing strategy by investigating repetition priming of compound words (e.g. buttercup) and their constituents (e.g. butter, cup). Eighty-eight right-handed participants (29 M, 59 F) completed a lexical decision experiment in which centrally-presented compounds primed related (whole compound, first constituent, second constituent) and unrelated targets presented laterally to the left or right visual field; participants made button-press word/nonword decisions. Consistent with the LH parallel/RH serial distinction, repetition priming prompted an RH advantage for first constituents, whereas the LH performed equally efficiently in response to both first and second constituents. These data thus highlight differences in the hemispheres’ language processing strategies, offering new evidence supporting a relative parallel/serial distinction in LH/RH visual word recognition.